Optical coherence tomography (OCT) is a non-contact, optical imaging modality that provides high-resolution cross-sectional images of the various layers of the anterior and posterior eye. In recent years, OCT has become the standard of care for diagnosing and monitoring therapy for many ophthalmic diseases, including age-related macular degeneration and diabetic retinopathy. OCT is also commonly used to aid in ophthalmic surgical planning and post-operative assessment, and more recently, has been used perioperatively via handheld probes mounted to the surgical microscope. Intrasurgical OCT systems integrated directly into the optical train of the surgical microscope are rapidly making their way to the clinic.
Each year, over 20 million ophthalmic surgeries are performed worldwide. Indications for ophthalmic surgery include potentially blinding diseases, such as cataracts, diabetic retinopathy, macular disease, and retinal detachment. Surgeons performing these delicate procedures are challenged by the translucent nature of tissues in the eye, making it nearly impossible to visualize microstructural changes during surgery. The high-resolution cross-sectional information provided by OCT is a natural complement to the microsurgical environment of an ophthalmic operating room. Intrasurgical OCT offers the surgeon the ability to see the microstructure of the eye in a way not possible with conventional surgical microscopes. By improving tissue visualization and providing surgical feedback, intraoperative OCT will enhance surgical precision, decrease surgical trauma, aid in surgical decision-making and ultimately improve functional and anatomical outcomes.
Preliminary research supports the utility of intrasurgical OCT and suggests that it may yield critical information regarding disease processes and the impact of surgical maneuvers, and thus aid surgical decision-making. Several surgical ophthalmic conditions have already been examined using intraoperative and perioperative imaging, including optic pit-related maculopathy, epiretinal membranes (ERM), macular holes, retinal detachments, and cataract surgery. Preliminary research has also shown that intrasurgical OCT can allow the surgeon to visualize the ultrastructural impact of a surgical maneuver on the tissue of interest. Documented changes in retinal architecture following ERM removal reveal alterations in retinal contour and, in some cases, microneurosensory retinal detachments. Intraoperative OCT during macular hole surgery has demonstrated changes in hole configuration following removal of the internal limiting membrane (ILM). Additionally, subclinical residual membranes have been identified that can be addressed during surgery. Finally, during intraocular lens (IOL) implantation, intraoperative OCT can be used to identify residual lens epithelial cells (LECs) in the posterior capsule, and also to evaluate adhesion of the posterior lens capsule to the IOL. Complete removal of residual LECs and good capsule-IOL adhesion are correlated with lower incidence of posterior capsule opacification, which occurs in as many as 30% of patients following cataract extraction.
Imaging of the ocular field in surgery benefits from a large depth of field in order to see the total physical extent of patho-physiology and trauma, to allow visualization of surgical instruments within the surgical field, and to allow the for the image OCT image to remain in view as the patient is subject to the various surgical manipulations that cause motion of the physical structure of the eye. Furthermore, the medium of the eye may become cloudy as the vitreous is stirred. For anterior imaging of the cornea and crystalline lens it is of specific interest to acquire deep images in order to visualize the entire affected optical structure during cornea and cataract surgeries. Surgical procedures require image depth and signal to noise improvement beyond the requirements of standard clinical diagnostic imaging.
In an article by Ruggeri et al entitled Imaging and full-length biometry of the eye during accommodation using spectral domain OCT with an optical switch, a reference arm switching technique is proposed to improve image depth in Fourier domain OCT (FDOCT), taking advantage of the nature of the Fourier domain signal processing. The signal to noise ratio (SNR) of Fourier domain images increases with distance from the path-matched, or direct current (DC) position of the image. Acquiring one image with the reference position on the proximal side of subject region and summing with a second image acquired with the reference position on the distal side of a subject region achieves the dual objectives of averaging to improve SNR generally, and balancing the SNR across the range of the image.
In order to be practically deployed, systems with reference arm switching function must operate in real-time, with image acquisition, image-pair rotation, registration, stitching and display occurring with minimum latency. These tasks require both rapid switching and fast and accurate registration algorithms.